Talking responsible AI use with University of Minnesota

November 30 marks the three year anniversary of ChatGPT — its launch changed the way everyday people access and interact with Generative AI (GenAI).  

People use GenAI for both academic and personal reasons — like grammar editing, note taking, meal planning and activity scheduling — to name just a few examples. Understanding how to be an effective and responsible user of GenAI is important not just for students, but for anyone who uses these tools. 

Lindsay Matts-Benson, teaching and learning program lead at University of Minnesota Libraries, answers questions about how you can be a better user of GenAI. 

Q: What are some ethical concerns adults may have when using AI in daily life? 
Matts-Benson: The balance between protecting privacy and sharing information with the tool is one concern. There is not a lot of transparency as to what is powering many of these tools and what information is collected about users. For example, there are tools that can take a photo and turn it into various cartoon-style images. But there’s not a lot of transparency around what is done with the uploaded photos and that could be an ethical consideration. Is using a cool tool worth losing privacy and agency over our own information? This also goes for GenAI tools for legal documents and health information. Adults may have concerns about the environmental impact of AI tools and the potential construction of data centers in their communities. I think adults are also aware of the potential malicious use of AI to create harmful viral or biased content. 

Q: What tips do you have for people to ensure they are using AI responsibly?
Matts-Benson: All users of AI should understand what GenAI tools learn from — what is in the data set. This data set doesn’t include all of the available information, but instead a curated collection of resources that can consciously exclude information. That can lead to one-sided or biased answers. GenAI output is just a game of probability — coming up with the most likely response based on what the system knows. Sometimes the responses make sense and other times they are like word salad. Most of all, I think folks need to consider when it is appropriate to use AI and what is okay to upload into a tool. AI tools aren’t the best for diagnosing medical issues or addressing mental health concerns. The best thing to do to be a responsible user of AI is to think about the why — why would I use AI for this, and why is AI the route to go?

Q: How can people identify if AI-generated information is true and reliable? 
Matts-Benson: While there are some things that are very clearly AI-generated, it is getting harder to discern what is real. I ask myself a few questions when I’m trying to figure out if something is true and reliable. How plausible is the thing that I’m reading or watching? Can I find this information in multiple places from identifiable sources? What information is backing up the claims in the content — or in librarian speak — is there evidence and citations? Is this a balanced or biased perspective on the topic? 

For videos, just remember that you know more about the real world than AI does. Do the things in the video move naturally? Does the background repeat itself on a loop? Are there nonsensical elements? Look for watermarks of AI creation tools on videos and watch for audio syncing issues. GenAI has a difficult time generating signs and text in images, so look for misspellings or words that don’t make sense. 

Q: What are the most common ethical dilemmas students run into when using GenAI? 
Matts-Benson: Students see this presumed efficiency of GenAI for completing certain coursework and have to strike the balance of completing assignments and the idea of experiencing the subject matter. This adds to the question of what learning experiences are appropriate for GenAI use. GenAI can augment the learning experience, but can’t learn things for you, and sometimes its responses are incorrect or inefficient. Students also have to find the balance of how much GenAI use is too much. What are the appropriate ways to acknowledge AI use while understanding the surrounding misconceptions? If GenAI use is required in a course for a project, can students object to using GenAI because of the environmental impact and ethical considerations?

Q: How does your work at the University of Minnesota help people become better citizens of information?
Matts-Benson: My work is focused on teaching core information literacy concepts — how we can find, use, evaluate and share information in a responsible way, not just with academic research, but in our everyday lives. I spend a lot of time helping undergraduate students learn core research skills. This summer I had the opportunity to lead the creation of GenAI+U: a tool to help enhance student understanding of GenAI tools. I really wanted to create something both thoughtful and useful that built on what students may already know. I wanted it to be relatable and accessible with both video and text components. It’s accessible to anyone, both as a learning experience and as an open educational resource for other schools. While it is built with undergraduate students in mind, I think it lays a strong foundation for critical AI literacy skills for anyone.  

Lindsay Matts-Benson is the teaching and learning program lead at the University of Minnesota Libraries in the Twin Cities. Lindsay's work is focused on leading the Libraries integration of information literacy into the student experience and leading the direction of the Libraries-wide Teaching and Learning Program. Lindsay is the creator of GenAI+U: A Student Learning Experience, a three-module set of foundational lessons on AI literacy.

About the University of Minnesota Libraries
The University of Minnesota Libraries is a strategic resource of the Twin Cities campus and also provides integral information system support for the University’s four campuses in Crookston, Duluth, Morris, and Rochester. Composed of 10 library facilities with collections of more than 8.1 million volumes — and with special collections valued at nearly $1 billion — the University of Minnesota Libraries has a history of strength in research collections and a longstanding record of contribution to resource sharing within the state and beyond. Learn more at lib.umn.edu.

About “Talking...with University of Minnesota”
“Talking...with University of Minnesota” is a resource whereby University of Minnesota experts answer questions on timely topics. News organizations are welcome to republish this content. If you would like to schedule an interview or have suggestions for future topics, please contact University Public Relations at [email protected].